When Google launched their broad match modified match type some years ago, some marketers hailed the demise of phrase match as something of a foregone conclusion.
Many, ourselves included, quickly shifted gears and rolled out new campaigns using a combination of broad match modified (BMM), exact match and negative match keywords. And others did not.
Let’s consider why we did this. First — and this should be writ large wherever AdWords advertisers congregate — no one likes broad match. No one except, perhaps, Google.
Using broad match is essentially lazy, the equivalent of holding out your hand with all your money it in and letting Google take what they want. They’re getting better — broad match is nowhere near as bad as it used to be — but it still allows Google a freedom you wouldn’t want your car mechanic or your electrician to have.
As a result, many advertisers instead opted for phrase match and exact match for their keywords, often having to develop hundreds of phrase match keywords to cover a multitude of possible search terms. For example, if we were trying to promote our blue widgets, we might start with “blue widgets” as a search term but soon have to add “widgets in blue,” “blue color widgets,” and so on.
With the introduction of BMM, advertisers were suddenly able to cover hundreds of possible search terms by simply using +blue +widgets. In our agency, we then monitor the Search Query Performance reports, pull out those search terms that deserve special attention and move them directly into exact match status, thus making the intermediate phrase match version redundant.
This streamlines the process and, with new build campaigns, seems to be the simplest way to get campaigns live and generating real data. We can then crunch this data faster to get to the kernels of optimization that will drive ongoing success.
Phrase match is still holding its own
We reached out to other marketers to discover how phrase match fits into their account management approach and found that reports of phrase match’s demise may be a little premature.
As Mark Kennedy from SEOM Interactive comments:
“While the changes to match types over the last few months have leant towards the demise of phrase match, it still has its place. I won’t remove it from older accounts where phrase match KWs have good history and conversion value (although I will prune those with no value). And if they are in accounts where ad groups are broken out by match type, I tend to keep them as well — again, especially where there is historical data. The other time I use them is when I am mining search query reports for converting phrases that I want to test. So I’ll use different match types on these new terms (including phrase match) to see what performs well. So while phrase match isn’t ‘necessary’ anymore we haven’t taken it totally out of the utility belt.”
Kennedy’s comments about historical value are particularly significant. Google loves historical data — especially when it comes to Quality Score, for example, and it could take a new BMM keyword some time to achieve a high QS already enjoyed by a long-term phrase match keyword that has consistently performed well.
While others echo these sentiments, Julie Friedman Bacchini of Neptune Moon feels that keyword matching is becoming increasingly fuzzier. She cites the changes in the keyword planner where match types are no longer segmented as an indication of the way Google sees match types. Coupled with the ending of the close variant opt out, the battle lines between marketers and the search engines are blurring. Interestingly, she concludes:
“It’s a little ironic, really, because it seems like now more than ever true phrase match would be helpful to advertisers. With matching getting fuzzier, the ability to zero in on a particular phrase or way of expressing a search term without having to sculpt it with negatives would be welcomed by many, I suspect!”
James Svoboda, partner at WebRanking, also agrees that the close variant rollout last year essentially made phrase match and Broad Match Modified the same with one significant exception: word order.
“Since we typically see similar results for search terms containing the same words, but in different word order, such as PPC Agency Minneapolis and Minneapolis PPC Agency, both of which can be targeted using one BMM Keyword: +Minneapolis +PPC +Agency, we’ve opted to simplify our account structure by using BMM instead of phrase match for most keywords. There are a few case-by-case situations that may arise where word order has a significant impact on performance, such “Used Cars” and “Cars Used” where the first one typically will be used by searchers with queries for preowned vehicles and the second one by searchers looking for cars used for specific tasks such as: cars used for off-roading. Even in these cases, we can use Negative Keywords to shape the queries that we are targeting through BMM Keywords.”
As a result, they have effectively removed phrase match from their arsenal and are looking forward to the day when both Google and Bing follow suit.
Whether that day will ever come is debatable. There are, perhaps, too many accounts that have been built and optimized over the years around a solid core of well-performing phrase match keywords. And, since phrase match isn’t broken, there may be no reason to try to fix it.
And for some of us, like Gil Hong from Seer, phrase match may simply be a habit that is hard to shake:
“I still use phrase match! Partly out of habit of how I build campaigns in Excel, and in some instances, phrase match still has a place for specific keyword intent.”
The consensus, therefore, seems to be that there is some life in phrase match after all. Where a well-performing phrase match keyword with great historical data and a good Quality Score is present, it could take some time for a Broad Match Modified keyword to catch up — and indeed, it would be hard to imagine the performance ever quite being matched. But where there are well-performing phrase match keywords, there are Exact Match versions willing and eager to usurp them.
So poor phrase match is getting pressure from both sides, and while it might be able to hold its own for the time being based on past glories, going forward, its days are still probably numbered. But then, we said that a couple of years ago!
Some opinions expressed in this article may be those of a guest author and not necessarily Search Engine Land. Staff authors are listed here.